The introduction of catalytic cracking to the petroleum industry in the 1930's constituted a major advance over previous techniques with the object of increasing the yield of gaseoline and its quality. Early fixed bed, moving bed, and fluid bed catalytic cracking FCC processes employed vacuum gas oils (VGO) from crude sources that were considered sweet and light. The terminology of sweet refers to low sulfur content and light refers to the amount of material boiling below approximately 1000-1025.degree. F.
The catalysts employed in early homogenous fluid dense beds were of an amorphous siliceous material, prepared synthetically or from naturally occurring materials activated by acid leaching. Tremendous strides were made in the 1950's in FCC technology in the areas of metallurgy, processing equipment, regeneration and new more-active and more stable amorphous catalysts. However, increasing demand with respect to quantity of gasoline and increased octane number requirements to satisfy the new high horsepower-high compression engines being promoted by the auto industry, put extreme pressure on the petroleum industry to increase FCC capacity and severity of operation.
A major breakthrough in FCC catalysts came in the early 1960's with the introduction of molecular sieves or zeolites. These materials were incorporated into the matrix of amorphous and/or amorphous/kaolin materials constituting the FCC catalysts of that time. These new zeolitic catalysts, containing a crystalline aluminosilicate zeolite in an amorphous or amorphous/kaolin matrix of silica, alumina, silica-alumina, kaolin, clay or the like were at least 1000-10,000 times more active for cracking hydrocarbons than the earlier amorphous or amorphous/kaolin containing silica-alumina catalysts. This introduction of zeolitic cracking catalysts revolutionized the fluid catalytic cracking process. Innovations were developed to handle these high activities, such as riser cracking, shortened contact times, new regeneration processes, new improved zeolitic catalyst developments, and the like.
The new catalyst developments revolved around the development of various zeolites such as synthetic types X and Y and naturally occurring faujasites; increased thermal-steam (hydrothermal) stability of zeolites through the inclusion of rare earth ions or ammonium ions via ion-exchange techniques; and the development of more attrition resistant matrices for supporting the zeolites. These zeolitic catalyst developments gave the petroleum industry the capability of greatly increasing throughput of feedstock with increased conversion and selectivity while employing the same units without expansion and without requiring new unit construction.
After the introduction of zeolite-containing catalysts the petroleum industry began to suffer from a lack of crude availability as to quantity and quality accompanied by increasing demand for gasoline with increasing octane values. The world crude supply picture changed dramatically in the late 1960's and early 1970's. From a surplus of light, sweet crudes the supply situation changed to a tighter supply with an ever-increasing amount of heavier crudes with higher sulfer contents. These heavier and higher sulfur crudes presented processing problems to the petroleum refiner in that these heavier crudes invariably also contained much higher metals and Conradson carbon values, with accompanying significantly increased asphaltic content.
Fractionation of the total crude to yield cat cracker charge stocks also required much better control to ensure that metals and Conradson carbon values were not carried overhead to contaminate the FCC charge stock.
The effects of heavy metals and Conradson carbon on a zeolite-containing FCC catalyst have been described in the literature as to their highly unfavorable effect in lowering catalyst activity and selectivity for gasoline production and their harmful effect on catalyst life.
These heavier crude oils also contained more of the heavier fractions and yielded a lower volume of the high quality FCC charge stocks which normally boil below about 1025.degree. F. and are usually processed so as to contain total metal levels below 1 ppm, preferably below 0.1 ppm, and Conradson carbon values substantially below 1.0.
With the increasing supply of heavier crudes, which yield less gasoline, and the increasing demand for liquid transportation fuels, the petroleum industry began a search for processes to utilize these heavier crudes in producing gasoline. Many of these processes have been described in the literature and include Gulf's Gulfining and Union Oil's Unifining processes for treating residuum, UOP's Aurabon process, Hydrocarbon Research's H-Oil process, Exxon's Flexicoking process to produce thermal gasoline and coke, H-Oil's Dynacracking and Phillip's Heavy Oil Cracking (HOC) processes. These processes utilize thermal cracking or hydrotreating following by FCC or hydrocracking operations to handle the higher content of metal contaminants (Ni-V-Fe-Cu-Na) and high Conradson carbon values of 5-15. Some of the drawbacks of these types of processing are as follows: coking yields thermally cracked gasoline which has a much lower octane value than cat cracked gasoline, and is unstable due to the production of gum from diolefins, and requires further hydrotreating and reforming to produce a high octane product; gas oil quality is degraded due to thermal reactions which produce a product containing refractory polynuclear aromatics and high Conradson carbon levels which are highly unsuitable for catalytic cracking; and hydrotreating requires expensive high pressure hydrogen, multi-reactor systems made of special alloys, costly operations, and a separate costly facility for the production of hydrogen.
To better understand the reasons why the industry has progressed along the processing schemes described, one must understand the known effects of contaminant metals (Ni-V-Fe-Cu-Na) and Conradson carbon on the zeolite-containing cracking catalysts and the operating parameters of an FCC unit. Metal content and Conradson carbon are two very effective restraints on the operation of an FCC unit and may even impose undesirable restraints on a Reduced Crude Conversion (RCC) unit from the standpoint of obtaining maximum conversion, selectivity and catalyst life. Relatively low levels of these contaminants are highly detrimental to an FCC unit. As metals and Conradson carbon levels are increased still further, the operating capacity and efficiency of an RCC unit may be adversely affected or made uneconomical. These adverse effects occur even through there is enough hydrogen in the feed to produce an ideal gasoline consisting of only toluene and isomeric pentenes (assuming a catalyst with such ideal selectivity could be devised).
The effect of increased Conradson carbon is to increase that portion of the feedstock converted to coke deposited on the catalyst. In typical VGO operations employing a zeolite-containing catalyst in an FCC unit, the amount of coke deposited on the catalyst averages around about 4-5 wt% of the feed. This coke production has been attributed to four different coking mechanisms, namely, contaminant coke from adverse reactions caused by metal deposits, catalytic coke caused by acid site cracking, entrained hydrocarbons resulting from pore structure adsorption and/or poor stripping, and Conradson carbon resulting from pyrolytic distillation of hydrocarbons in the conversion zone. There has been postulated two other sources of coke present in reduced crudes in addition to the four present in VGO. They are: (1) adsorbed and absorbed high boiling hydrocarbons which do not vaporize and cannot be removed by normally efficient stripping, and (2) high molecular weight nitrogen-containing hydrocarbon compounds absorbed on the catalyst's acid sites. Both of these two new types of coke producing phenomena add greatly to the complexity of resid processing. Therefore, in the processing of higher boiling fractions, e.g., reduced crudes, residual fractions, topped crude, and the like, the coke production based on feed is the summation of the four types present in VGO processing (the Conradson carbon value generally being much higher than for VGO), plus coke from the higher boiling unstrippable hydrocarbons and coke associated with the high boiling nitrogen-containing molecules which are adsorbed on the catalyst. Coke production on clean catalyst, when processing reduced crudes, may be estimated as approximately 4 wt% of the feed plus the Conradson carbon value of the heavy feedstock.
The coked catalyst is brought back to equilibrium activity by burning off the deactivating coke in a regeneration zone in the presence of air, and the regenerated catalyst is recycled back to the reaction zone. The heat generated during regeneration is removed by the catalyst and carried to the reaction zone for vaporization of the feed and to provide heat for the endothermic cracking reaction. The temperature in the regenerator is normally limited because of metallurgical limitations and the hydrothermal stability of the catalyst.
The hydrothermal stability of the zeolite-containing catalyst is determined by the temperature and steam partial pressure at which the zeolite begins to rapidly lose its crystalline structure to yield a low-activity amorphous material. The presence of steam is highly critical and is generated by the burning of adsorbed and absorbed (sorbed) carbonaceous material which has a significant hydrogen content (hydrogen to carbon atomic ratios generally greater than about 0.5). This carbonaceous material is principally the high-boiling sorbed hydrocarbons with boiling points as high as 1500.degree.-1700.degree. F. or above that have a modest hydrogen content and the high boiling nitrogen containing hydrocarbons, as well as related porphyrins and asphaltenes. The high molecular weight nitrogen compounds usually boil above 1025.degree. F. and may be either basic or acidic in nature. The basic nitrogen compounds may neutralize acid sites while those that are more acidic may be attracted to metal sites on the catalyst. The porphyrins and asphaltenes also generally boil above 1025.degree. F. and may contain elements other than carbon and hydrogen. As used in this specification, the term "heavy hydrocarbons" includes all carbon and hydrogen compounds that do not boil below about 1025.degree. F., regardless of the presence of other elements in the compound.
The heavy metals in the feed are generally present as porphyrins and/or asphaltenes. However, certain of these metals, particularly iron and copper, may be present as the free metal or as inorganic compounds resulting from either corrosion of process equipment or contaminants from other refining processes.
As the Conradson carbon value of the feedstock increases, coke production increases and this increased load will raise the regeneration temperature; thus the unit may be limited as to the amount of feed that can be processed because of its Conradson carbon content. Earlier VGO units operated with the regenerator at 1150.degree.-1250.degree. F. A new development in reduced crude processing, namely, Ashland Oil's "Reduced Crude Conversion Process", as described in pending U.S. applications Ser. Nos. 94,091, 94,092, 94,216, 94,217 and 94,227, all filed on Nov. 14, 1979, can operate at regenerator temperatures in the range of 1350.degree.-1400.degree. F. But even these higher regenerator temperatures place a limit on the Conradson carbon value of the feed at approximately 8, which represents about 12-13 wt% coke on the catalyst based on the weight of feed. This level is controlling unless considerable water is introduced to further control temperature, which addition is also practiced in Ashland's RCC processes.
The metal-containing fractions of reduced crudes contain Ni-V-Fe-Cu in the form of porphyrins and asphaltenes. These metal-containing hydrocarbons are deposited on the catalyst during processing and are cracked in the riser to deposit the metal or are carried over by the coked catalyst as the metallo-porphyrin or asphaltene and converted to the metal oxide during regeneration. The adverse effects of these metals as taught in the literature are to cause non-selective or degradative cracking and dehydrogenation to produce increased amounts of coke and light gases such as hydrogen, methane and ethane. These mechanisms adversely affect selectivity, resulting in poor yields and quality of gasoline and light cycle oil. The increased production of light gases, while impairing the yield and selectivity of the process, also puts an increased demand on gas compressor capacity. The increase in coke production, in addition to its negative impact on yield, also adversely affects catalyst activity-selectivity, greatly increases regenerator air demand and compressor capacity, and may result in uncontrollable and/or dangerous regenerator temperatures.
These problems of the prior art have been greatly minimized by the development at Ashland Oil, Inc., of its Reduced Crude Conversion (RCC) Processes described in the copending applications referenced above and incorporated herein by reference. The new processes can handle reduced crudes or crude oils containing high metals and Conradson carbon values previously not susceptible to direct processing.
It has long been known that reduced crudes with high nickel levels present serious problems as to catalyst deactivation at high metal on catalyst contents, e.g., 5000-10,000 ppm and elevated regenerator temperatures. It has now been recognized that when reduced crudes with high vanadium levels are processed over zeolite containing catalysts, especially at high vanadium levels on the catalyst, rapid deactivation of the zeolite can occur. This deactivation manifests itself as a loss of zeolitic structure. This loss has been observed at vanadium levels of 1000 ppm by weight or less. This loss of zeolitic structure becomes more rapid and severe with increasing levels of vanadium and at vanadium levels about 5000 ppm, particularly at levels approaching 10,000 ppm complete destruction of the zeolite may occur. Prior to the present invention, it was believed impossible to operate economically at vanadium levels higher than 10,000 ppm because of this phenomenon. Previously, deactivation of catalyst by vanadium at vanadium levels of less than 10,000 ppm has been retarded by lowering regenerator temperatures and increasing the addition rate of virgin catalyst. Lowering regenerator temperatures has the disadvantage of requiring higher catalyst to oil ratios which increase the amount of coke produced and adversely affect yields. Increasing catalyst addition rates is costly and can result in an uneconomical operation.
It has been found that vanadium is especially detrimental to catalyst life. The vanadium deposited on the catalyst under the reducing conditions in the riser is in an oxidation state less than +5. At the elevated temperatures and oxidizing conditions encountered in the regenerator the vanadium on the catalyst is converted to vanadium oxides, in particular vanadium pentoxide. The vanadium pentoxide has a melting point lower than temperatures encountered in the regeneration zone, and it can become a mobile liquid, flowing across the catalyst surface and plugging pores. This vanadia may also enter the zeolite structure, neutralizing the acid sites and, more significantly, irreversibly destroying the crystalline aluminosilicate structure and forming a less active amorphous material. In addition, this molten vanadia can, at high vanadia levels, especially for catalyst materials having a lower surface area, coat the catalyst microspheres and thereby coalesce particles which adversely affects their fluidization.